TEACHING SIMULATOR · SYNAPTIC PHYSIOLOGY
Low-calcium evoked end-plate potentials, after the classic experiments of Katz & Miledi and del Castillo & Katz. Each nerve stimulus releases a small, random number of transmitter quanta — including outright failures — and the amplitude of every resulting EPP is measured. Run enough trials and the histogram resolves into peaks spaced one quantum apart, revealing that release itself is a discrete, quantized process. Background: NCBI Bookshelf, NBK11028.
Each sweep represents one nerve stimulus delivered in a low-extracellular-calcium bath, which keeps the
average number of vesicles released per stimulus low enough that release failures are common — exactly the
condition Katz's group exploited. The number of quanta released on a given trial, n, is drawn
from a Poisson distribution with a fixed mean (the simulation holds [Ca²⁺] constant rather than making it
adjustable). The resulting EPP amplitude is n·q plus a small amount of trial-to-trial noise
(baseline recording noise, plus a little extra variability per quantum released), so each underlying peak
in the histogram has some width rather than being a perfect spike — as in the real recordings.
Because you can't read n directly off a single sweep (that's the whole problem the original
experimenters faced), the panel instead reports the same two classical estimators Katz's group used to
infer quantal content from purely macroscopic measurements: the failure method
(m = ln(N/N₀), from the fraction of silent trials) and the variance method
(m = mean²/variance, which also yields an independent estimate of the quantal size itself,
q = variance/mean, with no need to already know it). Toggle "reveal quantal content" to check
the true n against your read of the trace, and watch the two estimators converge on the true
values as more sweeps accumulate.
The quantal size itself is re-randomized every time the page loads (a tight range around the ~0.4 mV Boyd & Martin reported), so each student is measuring their own true value rather than already knowing the answer — reload the page for a fresh one. The "reveal" toggles show it as a check once you've made your own estimate, not before.
The ≈0 [Ca²⁺] button lowers mean quantal content to a small fraction of a quantum per stimulus, as in the original zero-calcium experiments: almost every sweep is a failure, and the rare successes are almost always exactly one quantum — this is what directly reveals the unitary EPP amplitude, since there's essentially no chance of two vesicles releasing on the same trial to confuse the reading. Note that the variance method for estimating m becomes unreliable in this regime (recording noise dominates the tiny signal variance); the failure method stays robust and is what the original experiments actually relied on here.